Full Data Sync Management

This tab, labeled Full Data Sync Management in the Cyber Vault side menu, is the central place to dispatch and browse vault copy job policies.

The Vault Jobs sub-tab lists all existing policies. You can run policies, select several, and start bulk full syncs.

Policies are created and edited in the Source Data Protection Group Config > Source Data Protection Policy Config (Step 5) wizard. This tab has no Add Config or Add Policy button.

Overview

A vault policy ties together:

• A source bucket
• A vault target bucket
• A schedule policy
• A protection group
• A set of credentials, or direct keys

Scheduled syncs in Vault Mode are run by the vault agent, and the protection group's schedule policy drives the dispatch. In Preparation Mode, you start incremental syncs by hand.

From this tab you can also start a full sync immediately for one or many policies. This is useful for initial seeding, target-skip migration passes, or re-baselining after a target wipe.

Configure the source and target S3 service accounts from the left navigation under Source Data Protection Group Config > Service Account Configuration. For target devices in Vault Mode, use the Vault Device Management page.

Vault Jobs sub-tab

The sub-tab lists all vault copy job policies. Each row is one copy pipeline from a source bucket (or bucket wildcard) to a vault bucket.

The table shows Name, Source Device, Namespace, Bucket, Tags, Protection Group, Mode, Schedule, Start, Window, Status, and Actions. The target settings below are part of the policy. Change them with Edit in Step 5.

Field	Description
Display Name	A human-readable label for the policy. It is shown in job history, the live queue, and the pending dashboard.
Source Device	The ObjectScale device that holds the source data.
Source Namespace	The S3 namespace on the source device.
Source Bucket	The bucket to copy, or * to include all buckets in the namespace that match the group pattern.
Target Device	The vault-side device that the policy writes to.
Target Namespace	The S3 namespace on the vault device.
Target Bucket Prefix	A string added in front of the source bucket name when the vault-side bucket is created or written to. It helps separate vault copies when several source devices write to the same vault namespace.
Bucket Group	The protection group the policy belongs to. It determines the target vault device, the schedule policy, and the operational mode.
Schedule Policy	Taken from the protection group. You cannot edit it per policy. Manage schedule policies on the Sync Schedules tab.
Schedule / Start / Window	The start hour and minute and the window duration of the assigned schedule policy.
Tags	Large and High Churn.
Mode	Prep or Vault.
Status	Enabled or Disabled.

A disabled policy is suspended, not deleted. It is skipped during scheduled dispatch and its Run button is disabled. Enable or disable a policy with Edit in Step 5.

Bulk action bar

An action bar appears at the top of the table when you check one or more rows.

Control	Description
Start Full Sync (N policies)	Opens the Bulk Sync modal. It is available even when Select all <N> is active across pages.
Refresh	Runs the policy list query against the server again.
Auto ON/OFF	Toggles whether the policy list loads automatically when the page opens. This helps on installations with thousands of policies.

info

Add, Delete, and Create Quick Policies are not on this tab. Use the Source Data Protection Group Config > Source Data Protection Policy Config (Step 5) wizard to add or remove policies.

Policies in Vault Mode

info

After Switch to Vault Mode, the vault agent runs the syncs. You cannot start or resume a full copy for those policies.

Run is disabled, and you can hover over it to see why. Start Full Sync skips those policies.

To run a full copy, click Exit Vault Mode on the Vault Data Sync Dashboard first.

One copy job per policy

A policy never has two copy jobs running at once. Two jobs would read the same source bucket and write the same target objects at the same time. That doubles the load and leaves extra object versions on a versioned or locked target.

A full copy of a bucket with very large objects can run for days. A daily Start Full Sync therefore often includes a policy whose previous copy is still running.

• That policy is skipped. The result lists it under "skipped — a copy job for it is still active", with the number of the job that is still running.
• The other selected policies are queued as usual.
• If every selected policy still has an active copy job, nothing is queued and the result says so.
• A policy selected twice in one request is queued once.

Run, Resume, Sync All, and Sync Group Now follow the same rule. Run and Resume answer "An active agent job already exists for this config".

A skipped policy is picked up by the next run after its active job finishes. You can also cancel the active job in Copy Jobs first.

Bulk Start Full Sync modal: Copy Mode

The Bulk Sync modal has a Copy Mode dropdown. It controls how the source bucket is reconciled against the target. The selection applies to every policy in the bulk dispatch.

Mode	Behavior
Standard	The default. Copies every source object regardless of target state. Use it for the first seed of an empty vault bucket.
Smart Skip	Reads the copy log in the database and skips objects that the console already copied in a previous successful run. It is the cheapest skip mode when this console wrote the target itself.
Target Skip	Compares the source bucket with the live target vault bucket and skips matches without writing. Use it to migrate from previous-product deployments, to restart after a half-finished copy, or whenever the target already holds some of the source data. See the Migration from Airgap help page for the full playbook.

Target Skip tiers

When you select Target Skip, the modal shows tier toggles.

Tier 1: strict ETag and size match

Always on. It is free and makes no extra requests.

Tier 1.5: date-trust skip

Optional and free. Select the box and enter a UTC cutoff date.

Source objects are skipped when both of these are true:

• Their LastModified is strictly before the cutoff date.
• Their size matches the target.

The cutoff is attested by the operator. If it is wrong, the risk of a false skip is yours. Use it only when nothing was written to the target bucket before the cutoff.

Tier 2: provenance markers

Optional. It adds one HEAD request per candidate object.

An object is skipped if the target object carries one of these markers:

• x-amz-meta-source-etag, written by every earlier copy from this console.
• x-amz-meta-x-emc-source-etag, the legacy marker written by ecs-sync.

Tier 2.5: range-hash content probe

Optional. It adds three ranged GETs on each side (source and target) per candidate object. It is a last-resort multipart skip when no cutoff date or sidecar is available.

• By default it samples 1 MB chunks and applies to objects of at least 5 MB.
• It detects changes only inside the sampled ranges. Small changes elsewhere in an object are likely missed.
• Use it only when the source bucket is read-only during the migration.

The defaults for Tier 2 and Tier 2.5 are configurable in Setup & Settings > Advanced Settings. They are pre-filled when you select Target Skip.

The Migration Playbook button in the modal opens the full migration guide.

Distribution plan modal

After you click Start Full Sync, the console fetches an anti-affinity distribution plan and shows it as a confirmation step. The plan has:

• A per-agent bar chart with job counts and percentage allocation. It covers only copy agents that are online and not draining. Copy jobs never run on the console.
• Heavy-bucket placements. Large or high-change-rate buckets are spread across engines so that no single engine is saturated.
• The total job count, split into heavy and standard jobs.

Confirm to submit the jobs, or cancel to back out without dispatching. The plan also appears on the per-group Sync Group Now button on the Vault Data Sync Dashboard.

Run on selection

Run on in the Run and Resume dialog offers Auto or a specific copy agent.

Situation	Result
Auto and no copy agent is available	The job is refused with "No copy agent available — start or undrain a copy agent and try again".
You pick a specific agent that is offline	The job is refused with "Copy agent "<name>" is offline — choose Auto or another copy agent, or start <name> and try again." This happens even when other agents are online.
The chosen agent was removed	The job is refused with "The selected copy agent no longer exists — choose Auto or another copy agent."

A job pinned to one agent runs only on that agent. If it is refused, choose Auto (any online agent) or another agent and run again. You can also start the agent's service and retry when it shows Online.

A refused Resume leaves the original job as it was, so you can resume it again.

Filter bar

The filter bar above the table narrows the policy list.

Filter	Description
Search	Free-text search on name, namespace, bucket, and device.
Device (All ECS / OBS Sources)	Filters to one source ObjectScale device.
Namespace (All Namespaces)	Filters to one namespace.
Bucket (All Buckets)	Filters to one bucket.
Group (All Protection Groups)	Filters to one protection group.
Mode	All modes (default), Prep, or Vault.

The Mode options mean:

• Prep shows policies in Preparation Mode protection groups, whose syncs are started from the console.
• Vault shows policies in Vault Mode groups, which the vault agent syncs on schedule.

All filters combine with AND. Select all <N> respects the Search, Device, Namespace, Group, and Mode filters, so bulk operations apply only to matching policies.

The Bucket filter only narrows the rows on screen, and Select all <N> ignores it. To act on one bucket's policies, use the checkbox in the table header instead.

Workflows

Suspend a policy temporarily

1. Find the policy in Source Data Protection Group Config > Source Data Protection Policy Config (Step 5).
2. Click Edit, set Enabled to Disabled, then click Save Policy.
3. The policy is skipped on the next scheduled dispatch. Running jobs for this policy are not affected.
4. To resume, enable it the same way. The Status column on Vault Jobs shows Enabled or Disabled.

Run a one-off full sync across many buckets

1. Filter the table to the scope you want (device, namespace, group, or mode).
2. Select rows. The header checkbox selects the current page. Select all <N> selects across all pages.
3. Click Start Full Sync in the action bar.
4. In the Bulk Sync modal, pick a Copy Mode (Standard, Smart Skip, or Target Skip) and any tier flags.
5. Confirm in the distribution plan modal.

Add new policies

Policy creation is in Source Data Protection Group Config > Source Data Protection Policy Config (Step 5). See that help page for:

• The bulk wizard, which assigns source and target accounts automatically.
• The Manual Policy form, which uses registered service accounts.

A policy appears on the Vault Jobs sub-tab as soon as you create it.

Tips

• Use * as the Source Bucket only when you want the policy to cover every bucket in the namespace automatically. For fine-grained control, create one policy per bucket in Step 5.
• Target Bucket Prefix avoids naming collisions when two source devices have buckets with the same name that write into the same vault namespace.
• Incremental sync windows are tracked per policy. Disabling a policy does not delete its changelist history. Re-enable it to resume from the last completed window.
• Target Skip Mode is read-only against the database. It never changes the console's stored state, so re-running a Target Skip pass is safe.
• Configure source and target S3 credentials before copy jobs can run. If Vault Jobs rows show "—" for the source or target account, configure them in Service Account Configuration.